You are on page 1of 48

Rock Mechanics 6, 189--236 (1974)

@ by Springer-Verlag 1974

Engineering Classification of Rock Masses


for the Design of Tunnel Support
By
N. Barton, R. Lien, and J. Lunde

With 8 Figures

(Received August 31, 1974)

Summary - - Zusammenfassung - - R6sum6

Engineering Classi[ication o[ Rock Masses [or the Design o[ Tunnel Support.


An analysis of some 200 tunnel case records has revealed a useful correlation be-
tween the amount and type of permanent support and the rock mass quality Q,
with respect to tunnel stability. The numerical value of Q ranges from 0.001 (for
exceptionally poor quality squeezing-ground) up to 1000 (for exceptionally good
quality rock which is practically unjointed). The rock mass quality Q is a function
of six parameters, each of which has a rating of importance, which can be estimated
from surface mapping and can be updated during subsequent excavation. The six
parameters are as follows; the RQD index, the number of joint sets, the rough-
ness of the weakest joints, the degree of alteration or filling along the weakest
joints, and two further parameters which account for the rock load and water
inflow. In combination these parameters represent the rock block-size, the inter-
block shear strength, and the active stress. The proposed classification is illustrated
by means of field examples and selected case records.
Detailed analysis of the rock mass quality and corresponding support practice
has shown that suitable permanent support can be estimated for the whole spec-
trum of rock qualities. This estimate is based on the rock mass quality Q, the
support pressure, and the dimensions and purpose of the excavation. The support
pressure appears to be a function of Q, the joint roughness, and the number of
joint sets. The latter two determine the dilatency and the degree of freedom of
the rock mass.
Detailed recommendations for support measures include various combinations
of shotcrete, bolting, and cast concrete arches together with the appropriate bolt
spacings and lengths, and the requisite thickness of shotcrete or concrete. The
boundary between self supporting tunnels and those requiring some form of per-
manent support can be determined from the rock mass quality Q.
Key words: Classification, rock mass, joints, shear strength, tunnels, support
pressure, shotcrete, bolts.
Rock Mechanics, Vol. 6/4 13
190 N. B a r t o n , R. Lien, and J. L u n d e :

Technische Klassifikation von GebirgsquaIitdt zwecks Projeletierens von HohI-


raumsicherungen im Fels. Eine Untersuchung yon Daten aus etwa 200 fertiggestellten
Tunnelbauten ergab einen nutzbaren Zusammenhang zwischen Umfang und Typ
des permanenten Verbanes und der Gebirgsqualit~,it Q. Die numerische Leitziffer
erfaf~t Werte yon 0,001 (iiut~erst schlechter, langsam rutschender oder quellender
Boden) bis auf 1000 fiir hochwertigen, fast bruchfreien Fels. Die Gebirgsqualit/it Q
ist eine Funktion yon sechs Parametern, die aus Oberfl;,ichenbeobachtungen und nach
skalierten Gewichten bestimmte Leitziffern erteilen werden. Die Werte k6nnen w~ih-
rend des Bauvortriebes justiert werden. Die sechs Parameter sin& RQD-Leitziffer,
Anzahl der Kluftsysteme, Rauhigkeit (fiir schw~ichste oder ungiinstigste Spaltebene),
Umwandlungsgrad (Charakter der Risse oder Fiillung liings der schw/ichsten Spal-
ten) und des weiteren zwei Parameter, die Spannungsniveau und Wasserzuflu~ be-
riicksichtigen. Wenn man diese Parameter koordiniert, vertreten sie den Einfluf~ der
K~rnung, der Scherfestigkeit an den Anschluf~fl~ichen zwischen den Felsbl6cken und
den einwirkenden Spannungen. Die vorgeschlagene Klassifikation wird mittels Bei-
spielen im Felde und einer Auswahl der Berichte aus fertiggestellten Anlagen er-
1/iutert.
Detaillierte Analysen der Gebirgsqualit~it und der entsprechenden Sicherungs-
mat~nahmen haben erwiesen, dat~ es m6glich ist, einen angemessenen Ausbau fiirs
ganze Spektrum der Gebirgsqualit~t zu veranschlagen. Die Bemessung ist auf die
Qualitiit Q des Gebirges, den Ausbaudruck und die Dimensionen und den Zweck
des Hohlraumes ausgerichtet. Der Ausbaudruck ist scheinbar eine Funktion yon Q
und v o n d e r Rauhigkeit und Anzahl der Spaltsysteme. Die beiden letzteren entschei-
den die Dilatanz der Felsmasse und den Freiheitsgrad der Felsbl6cke.
Detaillierte Anleitungen fiir Sicherungsmaf~nahmen umfassen verschiedene
Kombinationen yon N~igeln, Ankern, Spritzbeton und Ortsbetongew61ben sowie
auch Angaben iiber Ankerabst~inde und erforderliche St~irke des Spritz- oder Guf~-
betons. Die Grenze zwischen selbsttragenden Tunnels und denjenigen, die irgend
eine Art permanenten Verbaues ben6tigen, kann aus der Gebirgsqualit~it Q ermitteIt
werden.

Classification technique des roches en vue de I'dtude des sout~nements a prd-


voir dans Ies cavitds creusdes dans la roche. Une analyse de donnSes provenant de
quelque 200 cavities creus~es a permis d'&ablir une relation utile entre, d'une
part, l'envergure et le type de sout~nements permanents et, d'autre part, la qualit~
Q des masses rocheuses, en ce qui concerne la stabilitY. La valeur num~rique de
Q s'&end de 0,001 (roche particuli~rement mauvaise, fluante ou gonflant) jusqu'~
1000 pour une roche d'excelXente qualitY, pratiquement exempte de fissurations.
La qualit~ Q de la roche est une fonction de six param&res dont chacun, dans des
~chelles donn~es, s'est vu attribuer un coefficient pond&~ d~termin~ qu'on peut
esdmer en se basant sur des observations fakes en travaillant 5 ciel ouvert et qui
pourra &re ajust~ et mis ~ jour au cours de l'avancement des travaux. Ces para-
m~tres sont: l'indice RQD, le nombre de syst~mes de fissuration, la rugosit~ (celle du
plus faible plan de fissuration), le degr8 d'alt&ation (caract&istiques de ce dont
les fissures sont remplies), et, en outre, deux param&res qui tiennent compte du
niveau de tension et de l'afflux d'eau. Dans leur ensemble, ces param&res repr&
sentent l'influence qu'exercent la grandeur des pierres, la r&istance au cisaillement
existant sur les surfaces de contact entre les pierres, et les tensions actives. La clas-
sification sugg&~e est raise en &idence ~ l'aide d'exemples tir& de l'exp~rience
acquise sur le terrain ou tit& d'une s~lection de rapports concernant des ouvrages
ex&ut&.
Engineering Classification of Rock Masses for the Design of Tunnel Support 191

Des analyses ddtaill&s de la qualit6, accompagnde d'une prise en considdra-


tion de la pratique de southnement utilisde, ont permis de ddmontrer qu'il est
possible d'estimer un sout6nement approprid pour route la vari4td de qualitds de
roche. Cette estimation est basde sur la qualitd Q de la roche, sur la pression
supportde par le sout6nement, sur la taille de la cavitd et sur la destination de
celle-ci. La pression supportde par le sout6nement semble &re une fonction de Q
et de la rugositd et du nombre des syst6mes de fissuration. Ces deux derniers
param~tres semblent d&erminer la dilatance et le degr4 de libert6 (libertd de
mouvement) des pierres dans la roche.
Des recommandations d&aill&s de mesure de sfiret6 englobent diff&entes
combinaisons de b&on projet6, de boulonnage et d'arcs en b&on coulds, accom-
pagndes de l'indication de la distance appropride entre boulons, de la longueur de
ces derniers et de l'6paisseur 5 respecter tant pour le b6ton projetd que pour le
bdton could. La limite sdparant les cavit4s autoportantes de celles ndcessitant un
southnement permanent d'une mani~re ou d'une autre, peut &re d6terminde ~t
partir de la qualit4 Q de la roche.

Introduction
. . . . . when you can measure what you are speak-
ing about, and express it in numbers, you know
something about it, but when you cannot ex-
press it in numbers, your knowledge is of a
meagre and unsatisfactory kind . . . "
Lord Kelvin (1824--1907)
The Symposium on Large Permanent Underground Openings held in
Oslo in 1969 focussed attention on two important gaps in our ability to
design the correct support for excavations in rock masses. D en k h a u s (1970)
pointed out the existence of a missing link between the acquisition of rock
mechanics data and the final decisions as to whether an opening should be
lined, rock bolted, or kept unlined. B j e r r u m (1970) noted that the dilatent
property of many rock masses seemed to have been ignored when designing
rock bolt systems. He also doubted that the RQD index ( D e e r e , 1963) could
give a sufficiently complete description of a rock, since two rocks with the
same RQD index could show entirely different behaviour in a rock cavity.
The last criticism could also be levelled against other widely used rock
mechanics parameters, for instance: unconfined compressive strength, shear
strength, rock stress, joint frequency, etc. It is essential that such parameters
should each to be allowed to contribute in the final decision of tunnel support
requirements. The RQD index happens to be one of the better single pa-
rameters since it is a combined measure of joint frequency and degree of
alteration and discontinuity fillings, if these exist. However, it is relatively
insensitive to several important properties of rock masses, in particular the
friction angle of altered joint fillings ( C o r d i n g and D e e r e , 1972), and the
roughness or planarity of joint walls.
Despite the known limitations of RQD, several attempts have been
made to correlate it with the degree of tunnel support, as for instance by
13"
192 N. Barton, R. Lien, and J. L u n d e :

C e c i l (1970), D e e r e et al. (1970), and M e r r i t t (1972). In regularly jointed


and clay free rocks these attempts seem to be partly successful. However,
a one-parameter description of a rock mass is inevitably limited to a rela-
tively small number of geological environments, if it is to be reliable.
A more general method of numerically classifying rock masses and
estimating support has been described by W i c k h a m et al. (1972). This
includes a larger number of parameters, each having a numerical scale of
importance. B i e n i a w s k i (1973) has recently modified this system and
combined it with some other proposals for classification. The end result
is an eight-parameter description of jointed rock masses, each parameter
having five ratings of importance. The proposed parameters were: RQD,
degree of weathering, intact rock strength, spacing of joints, separation of
joints, continuity of joints, ground water inflow, strike and dip orientations.
In retrospect it would appear that both, W i c k h a m et al. (1972) and
B i e n i a w s k i (1973), have almost ignored three important properties of
rock masses, namely the roughness of joints, the frictional strength of joint
fillings, and the rock load.
The method of classifying rock masses to be described in this paper was
developed independently from that described by W i c k h a m et al. (1972)
and B i e n i a w s k i (1973). However, it is interesting to find that there are
several points in common. A special feature of the method is that it was
developed through exhaustive analysis of more than two hundred case
records. The recommendations for support are therefore detailed, and are
also based on estimates of support pressure, which can apparently be quite
closely estimated for the whole spectrum of rock mass environments.

Part I
Estimating the Rock Mass Quality

(A) D e v e l o p m e n t of t h e C l a s s i f i c a t i o n System

The tunnel case records described by C e c i l (1970) provided a com-


prehensive source of data for the initial development of the method. One of
Cecil's figures showed span width plotted against RQD for unsupported
tunnels. The trend for wider unsupported spans with higher RQD values
was recognizeable, although the scatter was large. The authors found that
this corrdation was improved if the relevant RQD values were divided by
a number representing the number of joint sets measured at each location.
As pointed out by C e c i l the number of joint sets is an important indication
of the degree of freedom of a rock mass.
The modified RQD had improved sensitivity to tunnel support require-
ments, since one important anomaly was removed. For example, a blocky
granitic rock mass having three joint sets and an RQD of 90 might give
equal tunnel stability to a tightly jointed phyllite, having only one joint set,
but an RQD of only 30.
Engineering Classification of Rock Masses for the Design of Tunnel Support 193

The importance of dilatancy and shear strength suggested further im-


provements to the modified R Q D . Joint roughness (small- and intermediate-
scale) was a potentially positive contribution to rock mass quality, while
joint alteration and filling materials were potentially negative. Two simple
numerical scales of joint roughness and alteration were therefore developed.
Finally, numerical scales for rock load and water pressure were added, to
further modify the original R Q D value.
Several months were spent in evaluating case records in the literature,
and developing improved numerical scales, until a consistent picture of
rock mass quality and tunnel support was obtained. Both the size of excava-
tion (span, diameter or height) and the purpose of the excavation (power
house, water tunnel, pilot heading, etc.) were additional important param-
eters for determining the type and degree of support. However, these two
parameters were not included in the estimation of rock mass quality. As
suggested by C o a t e s (1964), it is preferable that the estimate of rock mass
quality should be independent of both the type and size of excavation if it
is to be widely accepted as a classification system.

(B) M e t h o d for Estimating Rock Mass Quality Q

The six parameters chosen to describe the rock mass quality Q are
combined in the following way:

Q = ( R Q D / J n ) . (Jr/Jc~). (Jw/SRF) (1)


where
R Q D -- rock quality designation (Deere, 1963)
J,~ = joint set number
Jr = joint roughness number
J~ = joint alteration number
Jw = joint water reduction factor
SRF = stress reduction factor

The rock mass description and ratings for each of the six parameters
are given in Tables 1, 2 and 3. The range of possible Q values (approx.
0.001 to 1000) encompasses the whole spectrum of rock mass qualities from
heavy squeezing-ground right up to sound unjointed rock. (In fact more
than 300000 different geological combinations can theoretically be repre-
sented.) The case records examined included 13 igneous rock types, 24 meta-
morphic rock types, and 9 sedimentary rock types. More than 80 of the
case records involved clay mineral joint fillings of various kinds, including
12 swelling clay occurrences. However, most commonly the joints were
unfilled and the joint walls were unaltered or only slightly altered. Further
details of the range of case records studied can be found in the report by
B a r t o n et al. (1974).
194 N. Barton, R. L i e n , a n d J. L u n d e :

Table 1. D e s c r i p t i o n s and Ratings for the Parameters RQD, Jn, a n d Jr

1. ROCK QUALITY DESIGNATION (RQD)


A. Very p o o r . . . . . . . . . . . . . . . . . . . . . 0 - - 25 Note:
B. Poor . . . . . . . . . . . . . . . . . . . . . . . . . . 2 5 - - 50 (i) W h e r e RQD is reported or
C. Fair . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 0 - - 75 measured as < 10 (including
D. Good ......................... 7 5 - - 90 0) a n o m i n a l value of 10 is
E. Excellent . . . . . . . . . . . . . . . . . . . . . . 90--100 used to evaluate Q in Eq. (1)
(ii) RQD intervals of 5, i. e. 100,
95, 90, etc. are sufficiently
accurate
2. J O I N T SET N U M B E R (Jn)
A. Massive, no or few joints . . . . . . . . 0.5--1.0
B. One joint set . . . . . . . . . . . . . . . . . . . 2
C. One joint set plus r a n d o m . . . . . . . 3
D. T w o joint sets . . . . . . . . . . . . . . . . . 4
E. T w o joint sets plus r a n d o m . . . . . . 6
F. Three joint sets . . . . . . . . . . . . . . . . . 9
G. Three joint sets plus r a n d o m . . . . . 12 Note:
H. Four or m o r e joint sets, r a n d o m , (i) For intersections use
heavily jointed, "sugar cube", etc. 15 (3.0 x In)
J. Crushed rock, earthlike . . . . . . . . . . 20 (ii) For portals use
(2.0 x Jn)
3. JOINT ROUGHNESS NUMBER (Jr)
(a) Rock wall contact and
(b) Rock wall contact be[ore
10 cms shear
A. Discontinuous joints . . . . . . . . . . . . 4 Note:
B. R o u g h or irregular, undulating . . . 3 (i) A d d 1.0 if the m e a n spacing
C. Smooth, undulating . . . . . . . . . . . . . 2 of the relevant joint set is
D. Slickensided, undulating . . . . . . . . . 1.5 greater t h a n 3 m
E. R o u g h or irregular, p l a n a r . . . . . . . 1.5
F. Smooth, p l a n a r . . . . . . . . . . . . . . . . . 1.0 (ii) Jr=0.5 can be used for
G. Slickensided, planar . . . . . . . . . . . . . 0.5 p l a n a r slickensided joints
having lineations, provided
(c) No rock wall contact the lineations are favourably
when sheared orientated
H. Z o n e containing clay minerals thick
e n o u g h to prevent rock wall contact 1.0 (nominal)
J. Sandy, gravelly or crushed zone
thick e n o u g h to prevent rock wall
contact . . . . . . . . . . . . . . . . . . . . . . . . 1.0 (nominal)

T a b l e 2. D e s c r i p t i o n s and Ratings for the Parameters Ja a n d Jw

4. JOINT ALTERATION NUMBER (Ja) ~r (approx.)


(a) Rock wall contact
A. Tightly healed, hard, non-soften- 0.75 (--) Note:
ing, i m p e r m e a b l e filling i. e. (i) Values of (~0)r are in-
quartz or epidote tended as an approxi-
B. Unaltered joint walls, surface 1.0 (250__35 o) mate guide to the
staining only mineralogical proper-
ties of the alteration
C. Slightly altered joint walls. N o n - 2.0 (250--300) products, if present
softening mineral coatings, sandy
particles, clay-free disintegrated
rock etc.
D. Silty-, or sandy-clay coatings, small 3.0 (200--25 o)
clay-fraction (non-softening)
Engineering Classification of Rock Masses for the Design of T u n n e l Support 195

Table 2. Continued

E. Softening or low friction clay 4.0 (8°--160)


mineral coatings, i. e. kaolinite,
mica. Also chlorite, talc, gypsum
and graphite etc., and small
quantities of swelling clays.
(Discontinuous coatings, 1--2 mm
or less in thickness)
(b) Rock wall contact before
i0 cms shear
F. Sandy particles, clay-free dis- 4.0 (250--300)
integrated rock etc.
G. Strongly over-consolidated, non- 6.0 (16°--240 )
softening clay mineral fillings
(Continuous, < 5 mm in thickness)
H. Medium or low over-consolida- 8.0 (12°--160)
tion, softening, clay mineral
fillings. (Continuous, < 5 mm in
thickness)
J. Swelling clay fillings, i. e. mont- 8.0--12.0 (6°--120 )
morillonite (Continuous, < 5 mm
in thickness). Value of Ja depends
on percent of swelling clay-size
particles, and access to water etc.
(c) No rock wall contact
when sheared
K,L, Zones or bands of disintegrated 6.0, 8.0 (60--240 )
M. or crushed rock and clay (see G, or
H, J for description of clay con- 8.0--12.0
dition)
N. Zones or bands of silty- or sandy 5.0
clay, small clay fraction
(non-softening)
O,P, Thick, continuous zones or bands i0.0, 13.0 (60--240)
R. of clay (see G, H, J for &scrip- or
tion of clay condition) 13.0--20.0

5. JOINT WATER REDUCTION (Jw) Approx. water


FACTOR pressure
(kg/cm 2)

A. Dry excavations or minor inflow, 1.0 <1 Note:


i. e. < 5 l/rain, locally (i) Factors C to F are
B. Medium inflow or pressure 0.66 1.0-- 2.5 crude estimates. In-
occasional outwash of joint crease Jw if drainage
fillings measures are installed
C. Large inflow or high pressure in 0.5 2.5--10.0 (ii) Special problems
competent rock with unfilled caused by ice forma-
joints tion are not con-
D. Large inflow or high pressure, 0.33 2.5--10.0 sidered
considerable outwash of joint
fillings
E. Exceptionally high inflow or 0.2--0.1 > 10.0
water pressure at blasting, de-
caying with time
F. Exceptionally high inflow or O.1--0.05 > 10.0
water pressure continuing without
noticeable decay
196 N. B a r t o n , R. L i e n , and J. L u n d e :

Table 3. D e s c r i p t i o n s and R a t i n g s for the P a r a m e t e r SRF

6. STRESS REDUCTION FACTOR (SRF)


(a) Weakness zones intersecting excavation, Note:
which may cause loosening of rock mass (i) Reduce these values of
when tunnel is excavated SRF by 25--50% if the
relevant shear zones only
A. Multiple occurrences of weakness zones 10.0 influence but do not inter-
containing clay or chemically disintegrated sect the excavation
rock, very loose surrounding rock (any depth)
B. Single weakness zones containing clay, or 5.0
chemically disintegrated rock (depth of
excavation <50 m)
C. Single weakness zones containing clay, or 2.5
chemically disintegrated rock (depth of ex-
cavation >50 m)
D. Multiple shear zones in competent rock 7.5
(clay free), loose surrounding rock (any depth)
E. Single shear zones in competent rock (clay 5.0
free) (depth of excavation < 50 m)
F. Single shear zones in competent rock (clay 2.5
free) (depth of excavation >50 m)
G. Loose open joints, heavily jointed or "sugar 5.0
cube" etc. (any depth)
(b) Competent rock, rock stress problems

H. Low stress, neat- surface >200 >13 2.5 (it) For strongly anisotropic
J. Medium stress 200--10 13--0.66 1.0 stress field (if measured):
when 5<rrl/cra<i0, re-
K. High stress, very tight 10--5 0.66--0.33 0.5--2.0 duce ere and crt to 0.8 crc
structure (Usually
favourable to stability, and 0.8 err;
when ~ri/~a > 10, reduce %
may be unfavourable to and ¢t to 0.6 Crc and 0.6 et
wall stability) where: crc = unconfined
L. Mild rock burst 5--2.5 0.33--0.16 5--10 compression strength,
(massive rock) ¢~ = tensile strength
M. Heavy rock burst < 2.5 < 0.16 10--20 (point load), erl and era =
(massive rock) major and minor principal
stresses
(c) Squeezing rock; plastic flow of (iii) Few case records avail-
incompetent rock under the influence able where depth of crown
of high rock pressures below surface is less than
N. Mild squeezing rock pressure 5--10 span width. Suggest SRF
O. Heavy squeezing rock pressure 10--20 increase from 2.5 to 5 for
such cases (see H)
(d) Swelling rock; chemical swelling
activity depending on presence of water
P. Mild swelling rock pressure 5--10
R. Heavy swelling rock pressure 10--15

N o t e s on the Use o/ Tables 1, 2 a n d 3

W h e n m a k i n g e s t i m a t e s of t h e r o c k m a s s q u a l i t y (Q) t h e f o l l o w i n g guide-
lines s h o u l d b e f o l l o w e d , in a d d i t i o n to t h e n o t e s listed in T a b l e s 1, 2 a n d 3:
1. W h e n b o r e c o r e is u n a v a i l a b l e , R Q D can be e s t i m a t e d f r o m t h e
n u m b e r of joints p e r u n i t v o l u m e , in w h i c h t h e n u m b e r of joints p e r m e t r e
Engineering Classification of Rock Masses for the Design of Tunnel Support 197

for each joint set are added. A simple relation can be used to convert this
number to RQD for the case of clay-free rock masses ( P a l m s t r S m , 1974),

RQD = 115 --3.3 Jv'(approx.) (2)


where
]v - total number of joints per m 3
(RQD = 100 for ]~ < 4.5)
2. The parameter J, representing the number of joint sets will often
be affected by foliation, schistocity, slatey cleavage or bedding etc. If strongly
developed these parallel "joints" should obviously be counted as a complete
joint set. However, if there are few "joints" visible, or only occasional breaks
in bore core due to these features, then it will be more appropriate to count
them as "random joints" when evaluating J~ in Table 1.
3. The parameters J, and ](, (representing shear strength) should be
relevant to the weakest significant joint set or clay filled discontinuity in
a given zone. However, if the joint set or discontinuity with the minimum
value of (J,@,) is favourably orientated for stability, then a second, less
favourably orientated joint set or discontinuity may sometimes be of more
significance, and its higher value of (J,@,) should be used when evaluating
Q from Eq. (1).
4. When a rock mass contains clay, the factor SRF appropriate to
loosening loads should be evaluated (Table 3, 6 a). In such cases the strength
of the intact rock is of little interest. However, when jointing is minimal
and clay is completely absent, the strength of the intact rock may become
the weakest link, and the stability will then depend on the ratio rock-stress/
rock-strength (Table 3, 6b). A strongly anisotropic stress field is unfavour-
able to stability and is roughly accounted for as in note (ii), Table 3.
5. In general the compressive and tensile strengths (% and or) of the
intact rock should be evaluated in the direction that is unfavourable for
stability. This is especially important in the case of strongly anisotropic
rocks. In addition, the test samples should be saturated if this condition is
appropriate to present or future in situ conditions. A very conservative esti-
mate of strength should be made for those rocks that deteriorate when
exposed to moist or saturated conditions.
When the rock mass quality varies markedly from place to place it will
obviously be desirable to map and classify these zones separately. In general
the rock mass quality Q will be evaluated separately in two adjacent zones
if it is considered that a change in support will be justified in practice.
(A four-fold increase or reduction in Q, caused by a change in joint frequen-
cy, roughness or degree of alteration etc., will normally qualify for changed
support). However, if the variable zones intersect the excavations for only
a few metres, it will normally be most economical to map the overall quality,
and estimate a compromise value of Q, for eventual design of compromise
support. It is normally uneconomic to change support measures over very
short tunnel lengths, and in any case the overall stability has to be assured.
198 N. Barton, R. Lien, and J. L u n d e :

However, swelling and softening clay zones may often require individ-
ual sealing treatment, even if the affected discontinuities are quite narrow.
The type of treatment will depend on the clay content, the access to water,
and the quality of the wall rock ( S e l m e r - O l s e n , 1970). In some cases the
latter may be sufficiently high and the zone sufficiently narrow (i. e.
( 2 0 cms for it to be left unsealed. This will also depend on the use to
which the tunnel will be put. In general, individual classification and sealing
treatment for swelling or softening clay zones should be supplemented with
a compromise classification and support, so that the zones between the clay
are adequately supported.
Cases sometimes arise where unfavourably dipping shear zones delineate
exceptionally large unstable wedges requiring special support. This may
take the form of specially dimensioned tensioned anchors positioned to allow
for the variously orientated forces. A surge chamber wall at Churchill Falls
( B e n s o n et al., 1972) and a power house wall at Morrow Point ( B r o w n
et al., 1971) were both stabilized in this manner. In view of the speciab
nature of such problems, no attempt should be made to relate the relevant
rock mass quality Q to special-purpose support of this type.

(C) E x a m p l e s of R o c k M a s s Q u a l i t y Q
from Surface Exposures
Fig. 1 illustrates the method of classifying rock masses for their quality
Q. All the photographs are of surface exposures, but imaginary tunnel depths
of about 40 m have been assumed. Therefore, water pressures and rock pres-
sures of medium values have been assumed for each of the eight examples.
Beneath each photograph the following are listed:
1. Rock type.
2. Rock mass quality Q and values of the six parameters:
RQD/J,, Jr/J~, J~/SRF.
3. Numerical and alphabetical key to the classification descriptions
given in Tables 1, 2 and 3.
The classification of the six samples should be self explanatory. Each
numerical value can be checked against the relevant descriptions listed in
Tables 1, 2 and 3. The following list of observations may help to clarify
some of the special features of the method.
1. The positive contribution of irregular, undulating joints (Jr = 3) in
example 2, gives this heavily jointed rock mass almost the same quality (Q)
as example 1.

Fig. 1. Six examples of rock mass classified according to their tunnel stability
Sechs Beispiele yon Felsmassen, mit Riicksicht auf Tunnel-Stabilit~it klassifiziert
Six exemples de roches, classifides selon leur stabilitd dans le cas de cavitds creusdes
Engineering Classification of Rock Masses for the Design of Tunnel Support ]99

I. GRANITE 2. GRANITE
Q . 9 0 / g x l. 5/1.OxO. 66/1. O O ~ 70/15x3.0/1.0xO.66/1.0
10 (fair/good) = 9.2 (fair)
(IE/ZF, 3E/4B, 5B/6J) (lc/2H, 3B/4B, 5B/6J)

3. S A N D S T O N E - C L A Y S T O N E 4. NODULAR- LIMESTONE
Q - 40/9x 1.0/2.0x0.66/1.0 Q- 80/9xl. O/SxO.66/5
= 1.5 (poor) =0.24 (very poor)
(1B/2F, 3F/4C, 5B/6J) (1D/2F, 3J/4N, 5B/6G)

MUDSTONE (overall R Q D - 3 0 ) 6. GRANITE (decomposed) R Q D = 0


O ~ 3 0 / 9 x l . O/5xO.66/5 o- l o / z o x l . O/6x o. 66/6
= 0.09 (extremely poor) 0.009 (exceptionally poor)
(1B/ZF, 3J/4N, 5B/6B) (IA/2J, 3J/4K, 5B/6N)
200 N. Barton, R. Lien, and J. L u n d e :

2. The relatively widely spaced bedding joints in example 4 would nor-


mally produce a higher rock mass quality Q than for example 3. However,
the presence of layers of unconsolidated volcanic ash causes the rock mass
to be loose and unfavourable for tunnel stability.
3. The weakness zone in example 5 does not contain swelling or soft-
ening clay and therefore is not wide enough for individual classification.
The values of RQD, Jn, Jw and SRF are relevant to the overall rock mass
quality. However, the weakness zone does provide the minimum shear
strength parameters JflJa.
4. The decomposed granite shown in example 6 has a very low strength.
It is probable that at 40 metres depth, with a rock pressure in the region
of 10--15 kg/cm "a, the material will exhibit some mild squeezing, hence the
estimate of SRF = 6.

(D) S p e c i a l F e a t u r e s of t h e Six C l a s s i f i c a t i o n Parameters


Each of the parameter ratings listed in Tables 1, 2 and 3 are, with the
exception of RQD, the end product of successive modifications made during
analysis of the available case records. The successive modifications and
reanalyses were needed to improve the relation between the rock mass
quality Q and the support actually used. The final numerical ratings are
therefore more than just arbitrary descriptive scales such as poor (1), fair
(2), good (3) etc., and actually give some clue as to the principal properties
controlling tunnel stability in rock masses.
1. The first quotient appearing in Eq. (1) (RQD/Jn) represents the over-
all structure of the rock mass, and it happens to be a crude measure of the
relative block size, with the two extreme values (100/0.5 and 10/20) differ-
ing by a factor of 400. If as an example the quotient is interpreted in units
of centimeters, the extreme "particle sizes" of 200 cm and 0.5 cm are seen
to be crude but recogniseable approximations. Probably the largest block
should be several times this size and the smallest rock fragments less than
half the size. (Clay particles are of course excluded.)
2. The second quotient (J,./Ja) represents the roughness and degree of
alteration of the joint walls or filling materials. Quite by chance it was
found that the function tan -1 (JjJ~) is a fair approximation to the actual
shear strength that one might expect of the various combinations of wall
roughness and alteration products. Table 4 shows values of tan -1 (jjj~)o
tabulated for the three categories of rock wall contact given in Tables 1
and 2. It will be noticed that the "friction angles" are weighted in favour
of rough, unaltered joints in direct contact [category (a)]. It is to be expected
that such surfaces will be close to peak strength, that they will tend to dilate
strongly when sheared, and that they will therefore be especially favourable
to tunnel stability.
These high "friction angles" are very similar to the total friction angles
(combined cohesion and friction = tan -1 ~/~, where , = shear strength,
Engineering Classification of Rock Masses for the Design of Tunnel Support 201

o = n o r m a l stress), m e a s u r e d a n d p r e d i c t e d for such surfaces ( B a r t o n ,


1973). J o i n t s p a c i n g or block sizes larger t h a n 3 m will increase these esti-
m a t e s further, t h e r e b y a l l o w i n g for a possible scale effect [see n o t e 3(i),
T a b l e 1].
W h e n rock joints have t h i n clay m i n e r a l coatings a n d fillings [category
(b)], the s t r e n g t h is r e d u c e d significantly. Nevertheless, r e n e w e d rock wall

Table 4. E s t i m a t e of A p p a r e n t "Shear Strength" from the P a r a m e t e r s Jr and Ja

(a) Rock wall contact Jr tan -s (Yr/Ja) °


Ja = 0,75 1.0 2 3 4

A. Discontinuous joints 4 790 760 63o 530 45o


B. Rough, undulating 3 760 72o 56o 450 37o
C. Smooth, undulating 2 690 630 450 34o 270
D. Slickensided, undulating 1.5 630 560 370 270 210
E. Rough, planar 1.5 63o 56 o 37 o 270 210
F. Smooth, planar 1.0 530 450 27o 180 140
G. Slickensided, planar 0.5 34o 27o 140 9.5 ° 7.1 o

(b) Rock wall contact Jr tan -1 (~r/Ja) °


when sheared
Ja= 4 6 8 12

A. Discontinuous joints 4 45 o 34o 27o 180


B. Rough, undulating 3 37o 27o 210 14o
C. Smooth, undulating 2 27o 18o 140 9.5 o
D. Slickensided, undulating 1.5 210 14o 11o 7.1 o
E. Rough, planar 1.5 210 140 110 7.10
F. Smooth, planar 1.0 140 9.5 0 7.10 4.7 0
G. Slickensided, planar 0.5 7o 4.7 o 3.6o 2.4o

(c) No rock wall contact Jr tan -1 (Jr/Ja) °


when sheared
Ja=6 8 12

Disintegrated or crushed 1.0 9.5 o 7.1 o 4.7o


rock and clay

Bands of silty- or 1.0 Ja=5


sandy-clay
11o

Thick continuous bands 1.0 Ja = 10 13 20


of clay
5.7o 4.4o 2.9 o

c o n t a c t after small shear d i s p l a c e m e n t s have o c c u r r e d m a y be a very i m p o r -


t a n t f a c t o r for p r e s e r v i n g the e x c a v a t i o n s f r o m u l t i m a t e failure. T h e s e effects
have b e e n discussed b y B a r t o n (1974).
202 N. Barton, R. Lien, and J. L u n d e :

The third category involving no rock wall contact appears extremely


unfavourable to tunnel stability. The "friction angles" tabulated are a little
below residual strength values for most clays, and are possibly downgraded
by the fact that thick clay bands or fillings may tend to consolidate during
shear, at least if normally consolidated or if softening and swelling has
occurred. The swelling pressure of montmorillonite may also be a factor here.
3. The third quotation (J~/SRF) consists of two stress parameters. The
parameter J~ is a measure of water pressure, which has an adverse effect
on the shear strength of joints due to a reduction in effective normal stress.
Water may in addition cause softening and possible outwash in the case of
clay filled zoints. The parameter SRF is a measure of: (1) loosening load
in the case of excavation through shear zones and clay bearing rock, (2) rock
stress in competent rock, (3) squeezing or swelling loads in plastic incom-
petent rock. It can be regarded as a total stress parameter. It has proved
impossible to combine these two parameters in terms of inter-block effective
normal stress, because paradoxically a high value of effective normal stress
may sometimes signify less stable conditions than a low value, despite the
higher shear strength. The quotient (J~/SRF) is a complicated empirical fac-
tor describing the "active stresses".
It appears that the rock mass quality Q can therefore be considered
a function of only three parameters which are crude measures of:
1. block size (RQD/J,~)
2. inter-block shear strength (J,./Ja)
3. active stress (J~/SRF)
Undoubtedly, there are several other parameters which could be added to
improve the accuracy of the classification system. One of these would be
joint orientation. Although many case records included the necessary infor-
mation on structural orientation in relation to excavation axis, it was not
found to be the important general parameter that might be expected. The
parameters J,, J,. and J~ appear to play a more important general role than
orientation, because the number of joint sets determines the degree of free-
dom for block movement (if any), and the frictional and dilational charac-
teristics can vary more than the down-dip gravitational component of un-
favourably orientated joints. If joint orientation had been included, the clas-
sification system would be less general, and its essential simplicity lost.
However, it is recognised that orientation is an important parameter
in cases involving major clay-bearing weakness and fault zones. As suggested
earlier, it is not recommended that the classification system is extended to
cases involving special-purpose support, as would often be required in these
cases. Large unstable wedges, both underground and in rock slopes, require
specially orientated cable anchor or bolt systems. Special problems will in-
evitably require special classification systems. The six parameters chosen to
define the rock mass quality Q with respect to tunnel stability, will need
to be re-evaluated if the problem is one of drillability, boreability, ease of
excavation, slope stability etc. It seems very likely that the first four param-
Engineering Classification of Rock Masses for the Design of Tunnel Support 203

eters (RQD, J~, J,., J~) can form the basis for many rock mass classification
systems. However, the ratings may need to be modified, and other param-
eters added.

Part II
Estimating the Support Pressure

It is inevitable that all methods of tunnel excavation and support pre-


sently in use allow some degree of deformation in the surrounding rock.
In most of the poorer qualities of rock mass (squeezing and swelling rock
excluded), the final rock load tends to be greater if the initial support is
excessively soft (i. e. steel ribs and wooden blocking), or if the application
of support is delayed. The unchecked deformation may loosen a deeper zone
of rock above and around the excavation and the final loads will be greater
than they need be. The European approach using an immediate shotcrete
and/or rock bolt temporary support system therefore tends to minimise
final loads compared to rib and block methods, because it allows a con-
trolled amount of deformation sufficient to develop arching, but insufficient
to allow loosening.

(A) T e r z a g h i ' s E s t i m a t e s of S u p p o r t P r e s s u r e s
The support pressure criteria developed by T e r z a g h i (1946) were
mostly based on experiences in railway tunnels supported by steel ribs with
wooden blocking. For this reason his criteria tend to be over-conservative
in the better qualities of rock, if shotcrete and/or bolting is used as immediate
support in place of the steel ribs and wooden blocks. However, in the poor-
est qualities of rock it may be difficult to apply any type of support suffi-
ciently quickly to prevent significant deformation. As a result T e r z a g h i ' s
criteria appear quite relevant to present day practice when excavating me-
dium-size tunnels in very difficult rock conditions, and are in fact quite
widely used.
It is unlikely that a large range of tunnel sizes were involved in T er-
z a ghi's observations of the adequacy of support methods. Spans of between
5 and 10 m probably cover most of the tunnel sizes studied. It is therefore
appropriate in the first instance to consider his estimates of support pressure
relevant to this approximate size range. In Table 5 the support pressures
have been tabulated for each of the nine classes of rock mass loosely de-
fined by T e r z a g h i .
Although the accuracy of the above estimates of support pressures will
vary with the degree of deformation allowed, they do serve as a useful
guide as to the possible range that are likely to be encountered in practice.
Case records describing design pressures, or better still measured support
pressures, can be used to supplement and check these ranges. In each case
the support pressures will depend on both the rock mass quality and the type
of support method used.
204 N. B a r t o n , R. L i e n , and J. L u n d e :

Table 5. E s t i m a t e s of Roof S u p p o r t Pressures for T u n n e l s of 5rn and 10m


Span after T e r z a g h i (1946)
Assume: span = height, rock density 7 = 2.6 t/m a

Description Rock load estimates Support pressures kg/cm 2


(m) B=H=5m B=H=10m

1. Hard and intact zero 0 0


2. Hard stratified or schistose 0 to 0.5 B 0 to 0.6 0 to 1.3
3. Massive, moderately jointed 0 to 0.25 B 0 to 0.3 0 to 0.6
4. Moderately blocky and 0.25 B to 0.35 (B + H) 0.3 to 0.9 0.6 to 1.8
seamy
5. Very blocky and seamy (0.35 to 1.10) (B+H) 0.9 to 2.9 1.8 to 2.9
6. Completely crushed but 1.10 (B+H) 2.9 5.7
chemically intact
7. Squeezing rock, moderate (1.10 to 2.10) (B+H) 2.9 to 5.5 5.7 to 10.9
depth
8. Squeezing rock, great depth • (2.10 to 4.50) (B+H) 5.5 to 11.7 10.9 to 23.4
9. Swelling rock up to 80 m any (B+H) up to 20.8 up to 20.8

As a p r e l i m i n a r y effort to relate rock mass q u a l i t y Q to s u p p o r t pres-


sure, the a u t h o r s t r a n s l a t e d T e r z a g h i ' s n i n e rock mass descriptions into

Table 6. E s t i m a t e s of Rock Mass Q u a l i t y Q for the N i n e Classes of Rock


Mass Listed in T a b l e 5

No. RQD Jn Jr Ja Yw SRF Q (range)

1 100 --<2 4 1 1 1 >=200


2 =>30 3 1 1 1 1 20--10
3 100 6 > 1.5 1 1 1 50--25
4 80 9 1 <3 0.66 1 6--2
5 50 12 1 >3 0.66 1 1--0.4
6 20 15 1 2 =<0.66 5 0.08 --0.04
7 20 20 1 =)6 0.66 5--10 0.03 --0.01
8 0 20 1 >=6 0.33 10--20 0.004--0.001
9 0 20 1 12 < 0.66 10 0.003--0.001

values of the six classification p a r a m e t e r s , as s h o w n in T a b l e 6. T h e r e is


o b v i o u s l y r o o m for a l t e r n a t i v e i n t e r p r e t a t i o n . H o w e v e r , the resulting ranges
of Q were a useful s t a r t i n g p o i n t .

(B) E f f e c t of Dimensions

T h e r e is a f u r t h e r i m p o r t a n t factor which s h o u l d n o t be o v e r l o o k e d
w h e n a t t e m p t i n g to estimate the r e q u i r e d s u p p o r t pressure for a g i v e n
e x c a v a t i o n u s i n g T e r z a g h i's m e t h o d . T h i s c o n c e r n s e x c a v a t i o n d i m e n s i o n s .
Fig. 2, r e p r o d u c e d f r o m C o r d i n g et al. (1972), shows the s u p p o r t pressures
Engineering Classification of Rock Masses for the Design of Tunnel Support 205

actually designed for a number of large caverns excavated during the last
two decades or so. These case records are numbered in the figure as below:

1. Cavities I and II (NTS) 15. El Toro


2. Cavity II NTS (stabilized) 16. Norad
3. Cavity II NTS (at failure) 17. Tumut I
4. Poatina 18. Tumut II
5. Poatina (initial) 19. Tuloma
6. Poatina (final) 20. Outardes
7. Churchill Falls 21 Cruachan
8. Hoos 22. Vlanden
9. Harspranget 23. Northfield
10. Sackingen 24. Boundary
11. Hongrin 25. Ronco Val Grande
12. Morrow Point (upper half of wall)
13. Woh 26. Ronco Val Grande
14. Oroville (lower half of wall)

There does not appear to be any trend or necessity to increase the support
pressure with increasing dimensions o[ cavern. For the most part, roof
support pressures range from approximately 0.5 to 1.5 kg/cm 2, and wall
support pressures from approximately 0.3 to 0.7kg/cmh In general these
pressures are less than half the value they would be if T e r z a g h i ' s (1946)
design criteria had been rigidly followed.
Fig. 3 is a convenient illustration of this apparent discrepancy between
T e r z a g h i ' s design criteria and the support capacity currently designed for
large rock bolted caverns. Improvements in support methods over the years
are undoubtedly part of the reason for the discrepancy. However, it is
believed that the widely different dimensions are of equal or more im-
portance.
T e r z a g h i ' s (1946) recommendations were based on two types of obser-
vations; firstly on model arching experiments in sand which he compared to blocky
and seamy rock having "very large grains and little or no cohesion", and secondly
on observations of failure of the wooden blocks inserted behind the steel ribs in
various railway tunnels in the eastern Alps. It is unlikely that a large range of
tunnel sizes was involved in these in situ experiments, and it is unlikely that
T e r z a g h i ever intended his recommendations to be extrapolated to excavations
approaching one order of magnitude larger. It seems extremely unlikely that with
present-day support methods, doubling the span width would have the effect of
doubling the pressure on the supports, as implied by column 2 of Table 5, and by
the rock load factors (n) and (m) illustrated in Figs. 2 and 3. Provided the rock
around an excavation is held in place in a "closed ring" (using shotcrete if
necessary), the required support pressures should be more or less independent of
moderate increases dimensions, though strongly dependent on unchanged rock
Rock Mechanics, Vol. 6/4 14
206 N. Barton, R. Lien, and J. L u n d e :

E
v
• bolt
~ ~ ~ ~ ,:~,~ p
• anchor ~ p -'~ roof
= "C--~ =nB~
Dt
2"0

S
Cb

I,-
K
O
Q.
Q.
ill
1.0
"1~ Qls
13 ~ 1 7 /
O
at
20g 21~

10 20 30
EXCAVATION SPAN m.

• bolt P--
• anchor ~j_ .-mH~

i 2"0

1"0

e!
24i2~
22 ~g
10 20 30 40 50
EXCAVATION HEIGHT m.
Fig. 2. Design support pressures for the roofs and walls of some large caverns, after C o r d i n g
et a]. (1972). (;a represents the rock density)
Projektierter Ausbaudruck ffir Gew61be und W~inde in einigen Kavernen, laut u. a.
C o r d i n g 1972. (), = Raumgewicht des Felses)
Pression supportde par le sout~nement, pr~vue pour les vofites et parois de quelques grandes
cavit~s creusdes dans la roche, selon C o r d i n g et autres, 1972. (Y reprdsente la densit~ de
la roche)
Engineering Classification of Rock Masses for the Design of Tunnel Support 207

mass quality. Obviously the thickness of shotcrete or cast concrete arches needs
to be increased when, for a given rock mass quality, the dimensions are increased.
However, this does not necessarily imply an increase in support pressure. Bolt
spacing, though not bolt length, usually remains unchanged. In fact, the total load

V e r y poor Poor Fair Good


-Excell.

2.5
! Steel ribs and wood

\
! ex,ava.a.~ I
2.0
\
i

I!

0
I--
U 1.5
i,I.

Q
0
1.0
X
' % i
U
0 I ,arge
excavatlonsl !~ ,,
0.5

0.25

0.10
I (
25 50 75

ROCK QUALITY RQDX

Fig. 3. Comparison of roof support designs for steel rib supported tunnels (large displacements,
small excations) and for rock bolted caverns (small displacements, large excavations) after
Monsees (1970), and C o r d i n g and Deere (1972)
Vergleich zwischen projektiertem Gew61beausbau mit Stahlbogenst/itze (groi~e Ver-
schiebungen, kleine Querschnitte) und Ankerausbau (Heine Verschiebungen, grof~e Quer-
schnitte) laut Monsees 1970, C o r d i n g und Deere 1972
Comparaison entre, d'une part, le sout&nement de vofites prfivu sous forme d'arcs m&alliques
(grands d6placements, petites cavitds) et, d'autre part, sous forme de boulonnage (petits
ddplacements, grandes cavitds), selon Monsees 1970 et C o r d i n g et Deere 1972

capacity of the support system is increased, but not the pressure. It will therefore
be assumed that under most conditions found in practice, excavation dimensions
can be largely ignored where support pressures are concerned. This appears to be
in line with present-day practice.
208 N. Barton, R. Lien, and J. L u n d e :

oo o o ,,, ~ ~, oo
o

~-~fllli
?i//,~,~,;
~/ lIl *~IlI,IiH
//

~a
~oo
o

~.0

i
i
o o~
II

~0
>L <

"~ O'

U'l
<
>:, IE
~0 t,,;
aiO
i,~ I,. o

,g
o m ~, m

3~lrtSS3tld lUOddl'lS
Engineering Classification of Rock Masses for the Design of Tunnel Support 209

(C) R e l a t i o n s h i p between Support Pressure


and Rock Mass QualityQ
(i) Roo[ Support Pressure
An empirical equation relating permanent support pressure and rock
mass quality Q, which fits available case records quite well, was found to
be the following:

Proof= (~-r ) Q-1/a (3)


where
P, ooe = permanent roof support pressure in kg/cm 2
Jr = joint roughness number
Q = rock mass quality

The diagonal lines drawn in Fig. 4 and numbered with their respective Jr
values were plotted directly from this equation. The shaded envelope is the
authors' estimate of the range to be expected in practice according to avail-
able case records. The double dependence of support pressure on joint
roughness number J~ was deliberate and it appears to be realistic according
to available case records. The poorer qualities of rock mass are dominated
by more or less non-dilatent clay filled joints (Jr = 1.0 nominal), while the
better qualities tend to receive their improved Q values from the dilatent
property of interlocked non-planar rock joints. Accordingly, the shaded
envelope curves downwards, and for the very best qualities, drops below
J~ = 5 , which signifies discontinuous joints having a spacing of more than
3 metres.
It is not possible to introduce more variables in the chart shown in Fig. 4.
Nevertheless, Eq. (3) can be improved if the number of joint sets (joint struc-
ture number J,O is also included separately, besides its contribution to Q.
When the number of joint sets falls below three, the degree of freedom for
block movement is greatly reduced since three joint sets (or two plus ran-
dom) is the limiting case for three-dimensional rock blocks. The equation
below is offered as an improved version of Eq. (3).

P~oo~ 2Jn*/2(Q)-1/a
3 Jr (4)

Fig. 4. Empirical method for estimating the support pressure. Plotted points refer to case
records describing measured or designed roof support pressures. Case records for each of
the numbered points have been described by Barton et al. (1974)
Empirische Methode zur Berechnung des Ausbau-Druckes. Die eingetragenen Punkte beziehen
sich auf beschriebene Ffille, wo gemessene oder projektierte Ausbaudrucke angezeigt sind.
Die Daten der einzelnen, numerierten Punkte wurden u. a. yon Barton (1974) gesammelt
M6thode empirique pour le calcul de la pression support6e par le sout~nement. Les points
appliquds se r~f~rent ~ des cas d&rits, off lesdites pressions, mesur6es on projet~es, ont fit~
indiqu&s. Les donndes pour chacun des points num4rot6s ont dt6 r6unies par Barton et
autres 1974
210 N. Barton, R. Lien, and J. L u n d e :

Ud

,\\
-~ .l~\ o\.\ o
~8 ~ ~ ,~\, ,,v

I-.
o
~_I \ \ G.
Q.
~L \~\ U~

o ~\~ 0
o x

o ~,...o
I
L ~' I, x

L ~h \' \ \

II
g ~ cr
ii %
%
>,.
/,~ la
m
....I
,<

~ Cr

~n
\ :E

I \ ~J

'k
I

ii ~ %
\
\
I L \ \
US:l
(.uJ) LH91:I H JO'~l:ll:lWYIq 'NYd5 = NOISNIWIQ .I.NZlVAII'iO:I
Engineering Classification of Rock Masses for the Design of Tunnel Support 211

It should be noted that Eqs. (3) and (4) will give an identical estimate of
roof support pressure when the rock mass is intersected by three joint sets
(Jn = 9). Eq. (4) will give a lower estimate of support pressure than Eq. (3)
when there are less than three joint sets (no three-dimensional blocks), and
a higher estimate when there are more than three joint sets. This would
seem to be a realistic improvement, since it provides estimates of support
pressure largely in agreement with actual practice, and generally within the
shaded envelope in Fig. 4.

(ii) Wall Support Pressure


Several large excavations have been supported with almost equal pres-
sures on the walls and roof arch (see for instance B a r t h , 1972, concerning
Waldeck II). In other cases the wall pressure may be less than one third that
used in the roof arch. In the Churchill Falls power cavern ( B e n s o n et al.,
1972), the support pressures applied were approximately 1.3 and 0.4 kg/cm 2
in the roof and walls respectively, despite the presence of unfavourably
dipping foliation joints in the 45 m high walls. The trend towards higher
roof support pressure is shown unmistakeably in Fig. 2.
In view of the more favourable position of excavation walls as opposed
to roofs, it seems appropriate to consider a hypothetically increased "wall
quality" which will be some function of the general rock mass quality Q for
a given excavation. Analysis of case records to compare the permanent roof
and wall support pressures used in a given quality of rock mass provide the
necessary guidelines.
It is recommended that a hypothetical "wall quality" equal to 5 Q be
regarded as the maximum for use in the better qualities of rock mass when
Q ~ 10. (This may lead to a recommendation of zero support for the walls
of small excavations as shown later.) In intermediate qualities, i.e.
0.1 < Q <[ 10, in which the wall pressure is of more consequence, a value
of 2.5 Q could be used. In the worst qualities, i. e. Q < 0.1, where the wall
pressure (and floor pressure) can be almost equal to the vertical pressure,
a minimum value 1.0 Q should probably be used. In exceptional cases of
invert swelling due to water uptake, the floor and lower walls might require

Fig. 5. Tunnel support chart showing the box numbering for 38 categories of support. The
two plotted points refer to the worked example given on page 230
• = roof • = wall
Diagramm, welches die 38 Ausbaukategorien veranschaulicht. Die beiden markierten
Punkte beziehen sich auf das Arbeitsbeispiel, Seite 230
• = Gew61be [] = Wand
Graphique montrant 38 catdgories de sout~nements. Les deux points margu~s se r6f6rent
l'exemple de travail reproduit fi la page 230
• = vofite • --- paroi
212 N.Barton, R.Lien, and J. Lunde:

i~~wi '\'~ 'V\ V'~


\\ ¢::',',
I1111 iit~l

r~ 1111 o&
i : g
A
x

IN
II
;\o ~o\.o O
,'~ o .\
.... :. \\ ~ o O o't~°~, Itl
o~. ,,,~o ..~__,..' >-
I,-
O~ ~o ~ ~p .oo '~~" \\ ~ ,",,, e~~ \
\ III
III
? m
_J
<
-I
O
I/1
Ul
<

".o
~L
ill ..... i °;t°\~.\o
o
~ °~::~\ ~ ~ u
o

\
\

.=
too %Ol l i li i 1 ~
o

lllil t
\
o oo o o o . . . . .

llS]
('"') IHOIIH Jo']BHIaWVIQ 'NVdS = NOISN~IWIO IN]IVAInI:}:I
Engineering Classification of R o c k Masses for the Design of T u n n e l Support 213

a lower value of Q than used for the roof. These modified "wall qualities"
can be substituted directly in Eqs. (3) and (4), or read directly into the sup-
port pressure chart shown in Fig. 4.

Part III
Design of Support[Based on Case Records

(A) T u n n e l S u p p o r t C h a r t f o r A n a l y s i s of C a s e R e c o r d s

The method of classifying a rock mass for its quality Q was developed
by successive re-analysis of case records, until a consistent relationship was
obtained between Q, the excavation dimension, and the support actually
used. These three variables were inter-related by means of a support chart.
The final version of this chart is given in Fig. 5. It was arrived at after
several alterations and re-analyses of the case records. The box numbering
1 to 38 is used as a reference to the support category. Support measures that
are appropriate to each category are tabulated later.

Table 7. T h e Excavation S u p p o r t R a t i o (ESR) A p p r o p r i a t e to a V a r i e t y of


Underground Excavations

Type of excavation ESR No. of cases

A. Temporary mine openings etc . . . . . . . . . . . . . . . . . . . . . ca. 3--5? (2)


B. Vertical shafts: (i) circular section . . . . . . . . . . . . . . . . . ca. 2.5? (0)
(ii) rectangular/square section . . . . . . . ca. 2.0? (0)
C. Permanent mine openings, water tunnels for hydro
power (exclude high pressure penstocks), pilot tunnels,
drifts and headings for large excavations etc . . . . . . . . 1.6 (83)
D. Storage rooms, water treatment plants, minor road
and railway tunnels, surge chambers, access tunnels,
etc. (cylindrical caverns ?) . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 (25)
E. Power stations, major road and railway tunnels, civil
defence chambers, portals, intersections etc . . . . . . . . . . 1.0 (79)
F. Underground nuclear power stations, railway stations,
sports and public facilities, factories etc . . . . . . . . . . . . . ca. 0.8 ? (2)

The left-hand axis of the support chart gives the equivalent dimension
(De), w h i c h is a f u n c t i o n of t h e size a n d p u r p o s e o f t h e e x c a v a t i o n . T h e s p a n
o r d i a m e t e r a r e u s e d as d i m e n s i o n s w h e n a n a l y s i n g r o o f s u p p o r t , a n d t h e
d i a m e t e r o r h e i g h t f o r w a l l s u p p o r t . T h e e x c a v a t i o n s u p p o r t r a t i o (ESR)

Fig. 6. Support recommendations are based on the analyses of more than 200 case records.
Numbered case records are described by B a r t o n et al. (1974)
Die Ausbauanleitungen sind auf Untersuchung yon fiber 200 beschriebenen Anlagen
basiert. Numerierte F~ille sind u. a. yon B a r t o n (1974) beschrieben
Les recommandations pour le southnement se basent sur un examen de plus de 200 ouvrages
ddcrits. Les cas num6rot4s ont fit6 d6crits par B a r t o n et autres (1974)
214 N. Barton, R. Lien, and J. L u n d e :

which modifies these dimensions, reflects construction practice in that the


degree of safety and support demanded by an excavation is determined by
the purpose of the excavation, the presence of machinery, personell etc.
The list of ESR values given in Table 7 was developed through trial
and error as the most workable solution to the problem of variable support
practice. The number of case records relevant to each class of construction
are given in brackets. The degree of confidence in these figures will be
roughly in proportion to the number of relevant case records, hence the
question marks.

(B) E x a m p l e s of C a s e R e c o r d A n a l y s i s
More than two hundred case records were evaluated, and the relevant
values of Q and SPAN/ESR are plotted in Fig. 6. All the numbered points
refer to case records that are described in detail by B a r t o n et al. (1974).
In all, more than ninety of the case records were obtained from C e c i l
(1970), who visited and mapped a wide variety of tunnel conditions in
Scandinavia. In view of their importance to the development of the classi-
fication system, a selection have been reproduced in Tables 8 and 9, with
relevant sketches in Figs. 7 and 8. The case record numbering used by C e c i l
is unchanged.
The twelve case records have been chosen to illustrate a variety of
rock mass environments. The six-parameter classification (Tables 1, 2 and 3)
should be checked to verify the various ratings used. The values of rock
mass quality Q and SPAN/ESR are plotted in the tunnel support chart
(Fig. 6), and the relevant support category can be found from Fig. 5 (box
numbers 1 to 38 represent support category numbers).
The extreme right-hand columns of Tables 8 and 9 are termed "roof
support recommendation", and apart from category number, contain in
abbreviated form a description of the recommended roof support for the
given tunnel. This is based on the support used in all those case records
that plot within the same support category. A complete list of the recom-
mended support for each category is given in a later section (Tables 11, 12,
13 and 14).
In order to make support recommendations consistent and continuous be-
tween categories, some simple design concepts were used to rationalize the bolt
spacings and shotcrete or concrete arch thickness for each category. This compromise
solution was tailored to fit those case records giving detailed dimensions of bolt
patterns and shotcrete or concrete linings. It also supplied a reasonable estimate of
support dimensions for case records where the support was referred to in vague
terms, i.e. "systematic bolting and shotcrete". The simple design concepts for
rationalizing the support recommendations are given in an appendix at the end
of the paper.
(C) S e l f - S u p p o r t i n g Tunnels
The lower diagonal line of the tunnel support chart (Figs. 5 and 6) was
found to be the approximate boundary between self-supporting excavations,
Engineering Classification of Rock Masses for the Design of Tunnel Support 215

i /-- I

f l

- 12,5 m.

LEPTITE MYLONITE 24

tl , /~Z~CN>'/A\ y/A\'

QUARTZITE 39 SCHIST ~3

L,.5 m w i d t h : 6.5 m

SCHIST L8 GRA~, 26

Fig. 7. Sketches of the six case records described in Table 8, after Cecil (1970)
Skizzen der sechs F~lle, welche nach Cecil (1970) in Tabelle 8 beschrieben sind
Croquis repr~sentant les six cas dScrits dans le tableau 8, selon Cecil (1970)
216 N. B a r t o n , R. L i e n , and J. L u n d e :

o ~ ~o < ~ I~c~
t'-,

v
~i + d ~ I+ ~ u +

',D oct
b,: ,.1i
',.o
,q. ',.,o
,.-i
,-4

r~

o
o t--,i ".O
0,--~
,5
0 0

o 0,,.~
d
rq

o 0 0
0 o r-q
i~. x..O

O~
0
o~ 0
~ ' ~~ ' ~ "~2~
~ B

.N 0 ".~ ~ . ~ ~-~
©

o 0
0 0
",0 0

0
0",

0
,q
"CI

~.= o ~ IZ~.-a

-o
~V ~'~ ~ ~ ""
'-~o~

~o"u < i= o "~


~ :~z ° ~ o
cz~
,-.,id i~
~ •
~ 0
Engineering Classification of Rock Masses for the Design of Tunnel Support 217

o , ¢ . u.~

o ~
~ v
4-.o r.~
S

,,q IN
b,3
II
~D v.o

IN 8
b q~ O
d 0 [ c5
p.
U'3 t~

o. ~4 ~- 8

v~
o. ""
v--4

0 o

• G)

0 ¢; 0

,}
~ ~ z.~'~ ~ .~ ~
i O

I
t~

,.6 iI 8
~4

~.~ ~ £
~;> 9o
• ~ edg~
r.~ ",~, ]

:X'~ = ~4-a8 1i ~ -~ ~ _ =
0
~ "~.,.,.,.,.,.,.,.,.,.~
"o~ =~ ' ~L~'~ -o ~
~o~o~ ~ t~ 4- ~ ~ ,

~ v • -8-~ .-~ ~ s~ ~=
,~-~ ~
~ O ~ ¢) "~J
U G} ~.~ ~ ~~ ~
~,~ . ~ O ~ ,~

~ . ~ ~3 x z Na

e4 ~d
4.o

t~
2i8 N. Barton, R. Lien, and J. Lunde:

=o

0"~

OU..~
t'-.
0 0
v ~ II -I- ~ 1 1 ~+

p..

c~ m.

O~
r~
t~ o
0
0

cq cq
t~ 0
0

t~
0
I'--.

0
0 ~

0 0
0
v~

oo
0

t~ t~

~'~-'~ ~ ~'~ u,...~


~- 0.~ 0 ~-,
o ~.~= ~ ~ ~ •'0~.'.~
~ u ~~ ~ .~ ~~.-.'~
0 o .~ ~ / o •, ~ . ~ . ~ . ~ ~ .~ ~ , ~
~ ~ ~ • :0.,~
o -~ ~ u . ~ ..~ ~

m
~ o ~ ~ o o~ ~.~o.~ .
•~ ~.~= ~

~ ~ ~ I~'- 0 O0
~ . ~ ,

o
~~. ~
. =o~ ~ I ~ 0~" ~~: /
~ M M

0 0
~Z k.O
Engineering Classification of Rock Masses for the Design of Tunnel Support 219

• 0
~0 o
~Z
~H-F J rr

t'-.l O
t>- ~4
~4
v~ e~

r'-- 8
b- 0
~ O 0
eq

o O
0e4

0 o.
O~-~

r¢3
X
o O t'--1
0
"vl- o O 0

©
O
8
0
Z zO ~~
O

v~i
t~
eq ",Z,

0
e,I

£
S a:

~ " ~ " a e~ . ~ .~'~


,-~'~ N "8 ~ t~ ~
-~,-~ ~: o
~ o ~ o 0..~

,--~ ~ . . ~ o'~
~"d ~2TN
."~ ~
~ Z o~ O~ "~
O ~
-~ ~.~,~ ~- z~

,,~- u9 b-.
i---. b-.
220 N. B a r t o n , R. L i e n , and J. L u n d e :

~ C{ay seom
i open she~ting 'I joint~I partinl[ 7 sond- filled

/
w i d t h = 5.9 m
/

GRANITE 60 GRANITE 66

/ / //,'/I,#'I / I?~11

P l a n v i e w of tunnel

SCHISTOSE METAGREYWACKE 67 GRANITE 70

Plan view

N o overbreak ~ /
r-- Overbreak
/ / /

Piiar /..~ " ~ x / Plar 20 m


/ / / / / /
Sed~mentatiQn chambers / / _/ / 1

// / //~- Over breo.k

GRANITE 7A,75 GNEISS 7"7

Fig. 8. Sketches of the six case records described in Table 9, after Cecil (1970)
Skizzen der sechs Fiille, welche nach Cecil (1970) in Tabelle 9 beschrieben sind
Croquis repr6sentant les six cas ddcrits dans le tableau 9, selon Cecil (1970)
Engineering Classification of Rock Masses for the Design of Tunnel Support 221

and those requiring some form of permanent support. Classification data


for the thirty case records that plot on or below the boundary in Fig. 6 are
given in Table 10.
As can be seen from the table, out of the thirty case records there are
only six supported tunnels that plot below the boundary. The remaining one
hundred and ninety case records contain only six further exceptions to the
rule; i.e. unsupported tunnels that plot above the diagonal boundary. It
therefore appears that self-supporting tunnels can be predicted with accept-
able accuracy. The linear boundary can be approximated by the following
equation:
De' = 2 Q °.4 (5)
where
De' = limiting value of SPAN/ESR
Q = rock mass quality

The unsupported spans listed in Table 10 range from 1.2 to 100 metres.
Thus it does not appear that span-width need be a limiting factor, pro-
vided the rock mass quality is sufficiently high. In fact the Carlsbad lime-
stone caverns of New Mexico have unsupported spans of up to 190 metres,
presumably due both to the absence of joints and to a favourable stress field.
The classification data listed in Table 10 gives a good indication of the
"vital statistics" of self-supporting tunnels. It appears that a high RQD
value (mean RQD = 85 %) is common but not without exception. One joint
set is also a common characteristic, although the mean value of J~ is 2.9,
which represents one joint set plus random. None of these unsupported tun-
nels have more than three joint sets. In general the joints tend to be dis-
continuous or undulating (mean J,. = 2.6), though there are several examples
with smooth-planar joints. The two most important requisites appear to be
unaltered joints ( J ~ 1) and dry excavations (J,~ = 1). There are very few
exceptions to these two observations.

(D) T u n n e l Support Recommendations

Different personal, national and continental engineering practices lead


inevitably to variations in methods of support, even for the same quality
of rock. The majority of data has been obtained from European case records
due in particular to the ninety or so case records from Scandinavia (Cecil,
1970) and other Norwegian case records known to the authors. As a result
of this European-Scandinavian bias, and the belief that shotcrete and bolting
methods deserve most attention, many well documented case records have
been ignored. These include those describing steel rib support methods, free
span concrete arch roofs, and pre-cast sectional linings.
Small variations in support methods also occur in each category and
are due to rock mass differences, since a given value of Q is not unique,
but a combination of several variables. In order to separate the more ira-
Rock Mechanics, Vol. 6/4 iS
222 N. B a r t o n , R. L i e n , a n d J. L u n d e :

Table 10. C l a s s i f i c a t i o n Data for Self-Supporting Tunnels

Support Case Description of support RQD/Jn Jr/Ja Jw/SRF SPAN/ESR Q


category No. used (=De)

No. 0 6 none, S (1 app.) for pro- 60/2 2/1 1/1 9/1.6 60


tection from small stones
(no 8 none 70/2 1/1 1/1 9/1.6 35
support) 17 s b + S (1 app.) for protec- 100/2 1.5/1 1/1 9/1.6 75
tion from small stones
20 none 70/2 1/1 1/1 9/1.6 35
21 none 100/1 4/1 0.66/1 13/1.0 266
27 (near category 13) none 90/3 1/1 1/1 12.5/1.6 30
29 none 90/2 3/1 1/1 12.5/1.6 135
35 none 10/3 2/1 1/1 5/1.6 6.7
36 none 20/2 2/1 1/1 5/1.6 20
63 (near category 17) B 100/9 1/1 1/2.5 5.9/1.6 4.4
68 none 100/1/2 5/1 1/1 10/1.0 1000
70 none 40/2 1.5/1 1/2.5 8/1.6 12
74 (near category 9) none 100/2 1/1 1/1 12/1.3 16.7
77 (near category 5) sb 100/1 5/1 1/2.5 20/1.3 200
(50 bolts per 300 m)
78 none 90/2 1.5/1 1/2.5 5/1.3 27
87 none 100/1 4/1 1/1 11.25/1.6 400
91 none 90/2 1.5/1 1/1 12/1.3 67.5
96 none 100/1 4/1 71/2.5 15/1.3 160
101 b none 75/9 2/3 0.66/1 3.5/1.3 3.7
112 none 80/2 2/1 1/15 1.2/1.6 5.3
113 none 100/1 4/1 1/7.5 2.3/1.6 46
115 (near category 13) B 100/1 4/1 1/20 6.4/1.0 20
(1.0 m)
119b none 100/1 4/1 1/0.5 100/4 800
119c none 100/1 4/1 1/0.5 100/5 800
120a none 95/9 3/1 1/1 7/1.3 31.6
120b none 95/9 3/1 1/0.5 7/1.3 63
127a none or sb 100/4 3/1 1/0.5 20/5 75
127b none or sb 100/4 3/1 1/0.5 20/3 150
144 sb, 2 m long 90/4 1/4 1/1 3/1.3 5.6
150 none 100/4 2/1 0.5/0.5 6.1/1.3 50

Key: S = shotcrete (number of applications in brackets)


B = systematic bolting (mean spacing in brackets)
sb = spot bolting
Engineering Classification of Rock Masses for the Design of Tunnel Support 223

portant variations in support practice, the conditional factors RQD/Jn and


JflJ~ should be evaluated in addition to the overall quality Q. Two excava-
tions having the same rock mass quality Q, may in one case be bolted, and
in the other case only shotcreted. The conditional factor RQD/Jn describing
bloch size will normally separate these two cases. For instance, rock masses
with RQD/J,~ values larger than 10 will tend to be massive to blocky re-
quiring only bolting, while values less than 10 are likely to represent blocky
and jointed rock, which can often be adequately treated with shotcrete. In
other examples, the conditional factor J,./J¢ describing inter-block shear
strength may play a more important role. In some cases the equivalent
dimension (De) which is equal to SPAN/ESR can be used as a third con-
ditional factor to explain differences in support practice.
Tables 1l, 12, 13 and 14 contain the authors' recommendations for
permanent support for all 38 categories. It should be noted that the support
tables have been designed in the first instance to supply estimates of per-
manent roo[ support. Methods for estimating permanent wall support are
based on the hypothetical "wall quality" (range 1.0 Q to 5.0 Q) that was
discussed on p. 213. A complete worked example is given at the end of the
paper to illustrate the whole method.
It will have been noticed that no recommendations for temporary sup-
port have been discussed up to this point. Only a limited number of the
case records contained such details. Therefore any recommendation given
here will be an approximation, without the necessary back-up from case
records. Nevertheless in principle, a tunnel with given values of SPAN/ESR
and quality Q will obviously require reduced overall measures where tem-
porary support is concerned. Appropriate reductions in support can be
obtained by increasing the value of ESR to 1.5 ESR, and by increasing Q
to 5 Q. In other words, shifting a plotted point downward and to the right
hand side of Fig. 5, in the general direction of the NO SUPPORT boundary.
These factors would be applied equally to both the roof and wall, such that
any differences in roof and wall support would also be in operation for
temporary support.
It should finally be emphasised that the support recommendations con-
tained in this paper are based for the most part on general engineering prac-
tice for a given type of excavation. If for some reason the quality of drilling
and blasting is better or worse than that in normal practice, then the recom-
mended support will tend to be over-conservative or inadequate respectively.
However, there is an additional complication in that the appearance of the
excavated surfaces (apparent rock mass quality) tends to suggest either an
increased or a decreased Q value for these two cases. For instance, when the
drilling is poorly executed and hole alignment is bad, the degree of over-
break and need for support may increase considerably. Therefore, where
possible, the rock mass quality Q should be estimated from exposures ex-
cavated in a similar manner to that used in the final excavations. Where
this is not possible, allowances should be made, in particular with regard
to the value of J~ (joint set number) and to a lesser extent RQD.
15"
224 N. Barton, R. L i e n , and J. L u n d e :

Table 11. S u p p o r t M e a s u r e s f o r R o c k M a s s e s o f " E x c e p t i o n a l " , " E x t r e m e l y


G o o d " , " V e r y G o o d " , a n d " G o o d " Q u a l i t y (Q range: 1000--10)

Support Q Conditional factors P SPAN/ Type of Note


care- RQD/ SPAN/ kg/cm" ESR (m) s u p p o r t see
gory Jn Jr/Jn ESR(m) (approx.) p. 229

1" 1000--400 -- < 0.01 20--40 sb (utg) m

2* 1000--400 -- < 0.01 30--60 sb (utg) m

3* 1000--400 -- < 0.01 46--80 sb (utg)


4* 1000--400 -- <0.01 65--100 sb (utg) m

5* 400--100 -- 0.05 12--30 sb (utg)


6* 400--100 -- 0.05 19--45 sb (utg)
7* 400--100 -- 0.05 30--65 sb (utg)
8* 400--100 -- 0.05 48--88 sb (utg)
9 100--40 ->20 -- -- 0.25 8.5--19 sb (utg) --
<20 -- -- B (utg) 2.5--3 m --
10 100--40 >30 -- -- 0.25 14--30 B (utg 2 - - 3 m --
<30 -- -- B (utg) 1.5--2 m --
+ clm
11" 100--40 >30 -- -- 0.25 23--48 B (tg) 2 - - 3 m --
<30 -- -- B (tg) 1.5--2 m --
+ clm
12':" 100--40 >30 -- -- 0.25 40--72 B (tg) 2 - - 3 m --
<30 -- -- B (tg) 1.5--2 m --
+ clm
13 40--10 >10 >1.5 -- 0.5 5--14 sb (utg) I
>10 <1.5 -- B(utg) 1 . 5 - - 2 m I
<10 >1.5 -- B(utg) 1 . 5 - - 2 m I
<10 <1.5 -- B(utg) 1 . 5 - - 2 m I
+ S 2 - - 3 cm
14 40--10 >10 -- >15 0.5 9--23 B (tg) 1.5--2 m I, II
+ clm
<10 -- >15 B (tg) 1.5--2 m I, II
+ S (mr) 5 - - 1 0 cm
-- -- <15 B (utg) 1.5--2 m I, III
+ clm
15 40--10 > 10 -- -- 0.5 15--40 B (tg) 1.5--2 m I, II, IV
+ clm
<10 -- -- B (tg) 1.5--2 m I, II, IV
+ S (mr) 5 - - 1 0 c m
16" 40--10 > 15 -- -- 0.5 30--65 B (tg) 1.5--2 m I, V, VI
See +clm
note XII <15 -- -- B (tg) 1.5--2 m I, V, VI
+ S (mr) 1 0 - - 1 5 c m

* A u t h o r s ' estimates of support. Insufficient case records available for reliable estimation
of s u p p o r t requirements.
T h e type of s u p p o r t to be used in categories 1 to 8 will depend on the blasting technique.
S m o o t h wall blasting and t h o r o u g h b a r r i n g - d o w n may remove the need for support. R o u g h -
wall blasting m a y result in the need for single applications of shotcrete, especially where the
excavation height is > 25 m. Future case records should differentiate categories 1 to 8.
Key to Support Tables:
sb = spot bolting
B = systematic bolting
(utg) = untensioned, grouted
Engineering Classification of Rock Masses for the Design of Tunnel Support 225

T a b l e 12. S u p p o r t Measures for Rock Masses of "Fair" and "Poor" Quality


(Q r a n g e : 1 0 - - 1 )

Support Q Conditional factors P SPAN/ T y p e of s u p p o r t Note


cate- RQD/Jn Jr/Ja SPAN/ K g / c m 2 ESR (m) See
gory ESR (approx.) p. 229

17 10--4 >30 -- -- 1.0 3.5--9 sb (utg) I


=>10, < 3 0 - - -- B (utg) 1 - - 1 . 5 m I
<10 -- >6m B (utg) 1 - - 1 . 5 m I
+ S 2--3 cm
<10 -- <6m S 2--3 cm I
18 10--4 >5 -- >10 m 1.0 7--15 B (tg) 1 - - 1 . 5 m I, III
+clm
>5 -- < 10 m B (utg) 1 - - 1 . 5 m I
+ clm
< 5 -- _->10 m B (tg) 1 - - l . 5 m I, III
+S 2--3 cm
<5 -- < 10 m B (utg) 1 - - 1 . 5 m I
+ S 2 - - 3 cm
19 10--4 -- -- >20 m 1.0 12--29 B (tg) i - - 2 m I, II, IV
+ S (mr) 1 0 - - 1 5 c m
-- -- < 20 m B (tg) 1--1.5 m I , II
+ S (mr) 5 - - 1 0 c m
20* 10--4 -- -- >35 m 1.0 24--52 B (tg) 1 - - 2 m I, V, VI
See + S (mr) 2 0 - - 2 5 cm
note XII -- -- <35 m B (tg) 1 - - 2 m I, II, IV
+ S (mr) 1 0 - - 2 0 cm
21 4--1 >12.5 =<0.75 - - 1.5 2.1--6.5 B (utg) 1 m I
+ S 2--3 cm
< 12.5 =<0.75 - - S 2.5--5 cm I
-- >0.75 -- B(utg) lm I
22 4--1 >10, <30 >1.0 -- 1.5 4 . 5 - - 1 1 . 5 B (utg) i m + c l m I
<10 >1.0 -- S 2 . 5 - - 7 . 5 cm I
<30 =<1.0 -- B(utg) lm I
+ S (mr) 2 . 5 - - 5 c m
>30 -- -- B (utg) l m I
23 4--1 -- -- > 15 m 1.5 8--24 B (tg) 1 - - 1 . 5 m I, II, IV,
+ S (mr) 1 0 - - 1 5 c m VII
-- -- <15 m B (utg) 1 - - 1 . 5 m I
+ S (mr) 5 - - 1 0 m
24': 4---1 - - -- >30 m 1.5 18--46 B (tg) 1 - - 1 . 5 m I, V, VI
See + S (mr) 1 5 - - 3 0 c m
note XII -- -- < 30 m B (tg) 1 - - 1 . 5 m I, II, IV
+ S (mr) 1 0 - - 1 5 cm

* A u t h o r s ' e s t i m a t e s of s u p p o r t . Insufficient case r e c o r d s available f o r reliable e s t i m a -


tion of s u p p o r t r e q u i r e m e n t s .

(tg) = t e n s i o n e d , ( e x p a n d i n g shell t y p e f o r c o m p e t e n t r o c k m a s s e s , g r o u t e d p o s t - t e n s i o n e d
in very p o o r q u a l i t y r o c k m a s s e s ; see N o t e XI)
S = shotcrete
(mr) = m e s h r e i n f o r c e d
clm = c h a i n link m e s h
C C A = cast c o n c r e t e a r c h
(st) = steel r e i n f o r c e d
Bolt s p a c i n g s are given in m e t r e s (m). Shotcrete, or cast c o n c r e t e a r c h t h i c k n e s s is given
in c e n t i m e t r e s (cm).
226 N. Barton, R. Lien, and J. Lunde:

x~ >~ ~8 x >

I o m o ~ og
o

o o
~J r~ r~

v o v v
o~
v

o' ¢'q
oo
0
©
I
~4 ~0
?

>

lib
0 ¢'q r! ~4

~9
¢q eq r4

v
o o ~ o

0 t / I All v A V All Atl V

uo
~ O o
~9 A A VII ] I I J I I I

oo
A VII I I J I I I I I
~9

,q-
c5
I
o

o
I
c5
o
I ?
C~

b-.
x
k0
r-4
Engineering Classification of Rock Masses for the Design of Tunnel Support 227

x~ x~

× ×x ~ =;:>
L [ I

c,h

o ~.~

22 +E ,,-,.,~ v ,,.,.~ v v v

e'h ©
I I
r!.
r-q
T

0
o'h
0
o4
0
r~
8

E
0 0
t'q t'q

r I I P J J I I AlP V 8

r-,,i rq r-q

A A VII I I I I I I }

All u,h ©
u'b u,~

A VIP I All v A Vii V I }

I I
,5

<I
0
228 N. B a r t o n , R. L i e n , and J. L u n d e :

× ~ x
×

O~

0 CD 0 0
0
~,~i ~ F=~
oo i uo I
o o o u'~
~O TM on ~ O ~_~ i m 6 ~ rq

8
s+ ~ ~ + 4 ~ 8 +i ~ + < ÷ ~ + 6 + 88+££+
U~
0
ng
0
~4 eq
I
0
0
~4
4 0,
0

~O KO ~ t".l

~o
d 8
r~
0
~cS q
u~ ~lO

S~ ~S t~
t I I Ill All All v v All All v v
0
0 0
u
eh t ~

0 O 0 0
I I All All V 1 I I I 1 I I1 II
~0

2
0 All All V l I I l l l I II II ©

o o. o

o t
I I I Z
o ~.
O o d o o

0
.,x::
~4
~ a .<
I
Engineering Classification of Rock Masses for the Design of Tunnel Support 229

Supplementary Notes for Support Tables


I. For cases of heavy rock bursting or "popping", tensioned bolts with
enlarged bearing plates often used, with spacing of about l m (occa-
sionally down to 0.8 m). Final support when "popping" activity ceases.
II. Several bolt lengths often used in same excavation, i. e. 3, 5 and 7m.
III. Several bolt lengths often used in same excavation, i. e. 2, 3 and 4 m.
IV. Tensioned cable anchors often used to supplement bolt support pres-
sures. Typical spacing 2--4 m.
V. Several bolt lengths often used in some excavations, i.e. 6, 8 and 10 m.
VI. Tensioned cable anchors often used to supplement bolt support pres-
sures. Typical spacing 4 - - 6 m.
VII. Several of the older generation power stations in this category employ
systematic or spot bolting with areas of chain link mesh, and a free
span concrete arch roof (25--40 cm) as permanent support.
VIII. Cases involving swelling, for instance montmorillonite clay (with access
of water). Room for expansion behind the support is used in cases
of heavy swelling. Drainage measures are used where possible.
IX. Cases not involving swelling clay or squeezing rock.
X. Cases involving squeezing rock. Heavy rigid support its generally used
as permanent support.
XI. According to the authors' experience, in cases of swelling or squeezing,
the temporary support required before concrete (or shotcrete) arches
are formed may consist of bolting (tensioned shell-expansion type) if
the value of RQD/J~ is sufficiently high (i. e. ~ 1.5), possibly com-
bined with shotcrete. If the rock mass is very heavily jointed or crushed
(i. e. RQD/J~ ~ 1.5, for example a "sugar cube" shear zone in quartz-
ite), then the temporary support may consist of up to several applica-
tions of shotcrete. Systematic bolting (tensioned) may be added after
casting the concrete (or shotcrete) arch to reduce the uneven loading
on the concrete, but it may not be effective when RQD/J,~~ 1.5, or
when a lot of clay is present, unless the bolts are grouted before ten-
sioning. A sufficient length of anchored bolt might also be obtained
using quick setting resin anchors in these extremely poor quality rock-
masses. Serious occurrences of swelling and/or squeezing rock may
require that the concrete arches are taken right up to the face, pos-
sibly using a shield as temporary shuttering. Temporary support of
the working face may also be required in these cases.
XII. For reasons of safety the multiple drift method will often be needed
during excavation and supporting of roof arch. Categories 16, 20, 24,
28, 32, 35 (SPAN/ESR ~ 15 m only).
XIII. Multiple drift method usually needed during excavation and support
of arch, walls and floor in cases of heavy squeezing. Category 38
(SPAN/ESR ~ 10 m only).
230 N. B a r t o n , R. Lien, and J. L u n d e :

(E) W o r k e d E x a m p l e
2 0 m S p a n M a c h i n e H a l l in P h y l l i t e

(i) Rock Mass Classification

Joint set 1 strongly developed foliation


smooth-planar (J,. = 1.0)
chlorite coatings (Ja = 4.0)
ca. 15 joints/metre

Joint set 2 smooth-undulating (J,. = 2)


slightly altered joint walls (]c, = 2)
ca. 5 joints/metre
J~=15+5=2o RQD=50 (Eq. 2)
J~ = 4. Minimum Jr/Ja = */4
Minor water inflows: J~,~= 1.0
Unconfined compression strength of phyllite (Oc) = 400 kg/cm e
Major principal stress (ol) = 30 kg/cm 2
Minor principal stress (oa) = 10 kg/cm 2
(ol/o3) = 3
Oc/q = 13.3 (medium stress) SRE = 1.0

50 1 1
Q 4 4 1 =3.1 (poor) (Eq. 1)

(ii) Support Recommendation

Type of excavation Machine hall B = 20 m, H = 30 m


(ESR = 1.0) B/ESR = 20, H/ESR = 30

Support category (a) Roof Q = 3.1; category 23 (Fig. 5)


(b) Walls " Q " = 3.1 • 2.5; category 20

Recommended Support (a) Category 23 Table 12 B(tg) 1.4m (Roof)


+ S (mr) 15 cm
Notes: II, IV, VII
(b) Category 20 Table 12 B(tg) 1.7m (Walls)
+ S (mr) 10 cm
Notes: II, IV
Engineering Classification of Rock Masses for the Design of Tunnel Support 231

Mean length
of bolts and anchors (a) Roof bolts 5.0 m (Appendix)
anchors 8.0 m (Appendix)
(b) Walls bolts 6.5 m (Appendix)
anchors 10.5 m (Appendix)
Support pressure estimates
(a) Roof Q=3.1 1. (Fig. 4, shaded envelope)
Approx. range for P,-oof= 0.9--2.0 kg/cm e
2. (Eq. 3) Proof = 1.37 kg/cm "~
3. (Eq. 4) Proof = 0.91 kg/cm ~
(b) Walls "Q" = 3.1 • 2.5 1. (Fig. 4, shaded envelope)
Approx. range for Pwall = 0.6--1.4 kg/cm e
2. (Eq. 3) P,,m = 1.01 kg/cm e
3. (Eq. 4) P,,.~i1= 0.67 kg/cm e

Commentary
1. Note the use of the minimum value J,./J(t for calculating Q. The prop-
erties of the joint set having the lowest shear strength should always be
used, unless the user considers the orientation is entirely favourable such
that a second joint is more unfavourable to stability, despite having a higher
value of J,./Ja.
2. The choice of 1.4 m and 1.7 m spacing for roof and wall bolts from
the empirical listed ranges of 1--1.5 m and 1--2 m was made in accordance
with the specific value of Q, in relation to the range for the given category
(i. e. Q = 1--4). These bolt spacings are approximate and need to be checked
against required support pressures.
3. When using Tables 11, 12, 13 and 14 for wall support, the relevant
span should be used when the conditional factor (SPAN/ESR) is listed. Hence
the choice of the minimum 10 cm of mesh reinforced shotcrete from a pos-
sible range of 10--20 cm.
4. The mean bolt and anchor lengths should be coordinated with the
recommendation given under Note II (p. 229). Thus, for the roof, variable
(intermeshed) bolt lengths of 3, 5 and 7m appear reasonable, while for the
wall 5, 6.5 and 8 m might be more appropriate. The recommendation for
using tensioned cable anchors (Note IV) is based on current practice in most
caverns of this size. The effectiveness of such widely spaced (2--4 m) rein-
forcement is perhaps open to question.
5. The range of estimates of support pressure give room for choice. The
estimates obtained from Eq. (4) are especially dependent on the absence of
additional joint sets. Should some additional random joints be discovered
when access tunnels are driven into this hypothetical rock mass, both J,, and
Q will be affected, and this will have a multiple effect on Eq. (4). The value
of J~ will increase to 6, Q will reduce to 2.1, and the estimate of roof sup-
port pressure would rize from 0.91 to 1.28 kg/cm 2.
232 N. Barton, R. Lien, and J. L u n d e :

Appendix
Design Concepts for Rationalizing the Support Tables
The simple theory used to rationalize the support dimensioning can be
conveniently divided into three parts: bolting, concrete lining, shotcrete
lining.

1. B o l t i n g
The support pressure capacity of tensioned or grouted bolts is equal
to the yield capacity of one bolt (if adequately anchored) divided by the
square of the bolt spacing. If a 10 tons working load is assumed for a
20 mm diameter bolt, the support pressure is as follows:
p = 1/a 2 (6)
where
P = support pressure capacity in kg/cm 2
a = bolt spacing in metres

Eq. (6) and the support pressure chart (Fig. 4) were used in combination with
the case records, and this helped to provide a rational and reasonably con-
tinuous spectrum of bolt spacings. When a range of spacings is quoted in
Tables 11, 12, 13 and 14, for instance 1.5 to 2.0 m, the lower limit applies
to the lowest rock mass quality Q, and the upper limit to the highest rock
mass quality in each given support category. In cases where anchors were
noted as a supplementary reinforcement method, the given bolt spacings
could be increased, provided the total support pressure generated by the
combined bolting and anchoring was not reduced.
Bolt and anchor lengths depend on the dimensions of the excavations.
Lengths used in the roof arch are usually related to the span, while lengths
used in the walls are usually related to the height of the excavations. The
ratio of bolt length to span tends to reduce as the span increases. This
trend has been demonstrated by B e n s o n et al. (1971). Accordingly, the
following recommendations are given as a simple rule-of-thumb, to be
modified as in situ conditions demand.

Roof: bolts L = 2 + 0.15 B/ESR (7)


anchors L = 0.40 B/ESR (8)

Walls: bolts L = 2 + 0.15 H/ESR (9)


anchors L = 0.35 H/ESR (10)
where
L = length in metres
B = span in metres
H = excavation height in metres
ESR = excavation support ratio
Engineering Classification of Rock Masses for the Design of Tunnel Support 233

2. C o n c r e t e L i n i n g

The theory of thin walled cylinders provides a simple expression be-


tween lining thickness, resultant stress in lining, and uniform internal or
external pressure at equilibrium. For external loading the following expres-
sion is obtained:

t- P'n (11)

where
P - externally applied pressure (kg/cm "2)
o = compressive stress in lining (kg/cm 2)
R = internal radius of lining (cm)
t = wall thickness for equilibrium (cm)

The above expression is based on the assumption that bending and shear
stresses are absent.
When a concrete lining is used in combination with systematic bolting,
stresses caused by uneven loading or non-circular linings can presumably
be minimized and the above equation used with a conservative value for
allowable stress. If bolt tensions could be guaranteed, some sharing of sup-
port pressure would occur and lining thickness could be reduced. However,
some form of internal steel reinforcement may be required to reduce the
unfavourable effect of uneven stresses. A conservative value of ~ (allowable)
equal to 50 kg/cm 2 was assumed when rationalizing Tables 11, 12, 13 and 14.
The appropriate range of pressure (P) was estimated using Fig. 4, in com-
bination with available case records.
Support pressure load sharing by systematic bolting was ignored, there-
fore concrete thickness may be too conservative if bolts are added and an-
chorage is effective. However, it should be emphasised that concrete lining
is only recommended in the poorest qualities of rock mass, where the effec-
tiveness of bolt and~orage is relatively uncertain.

3. S h o t c r e t e L i n i n g

When single (2--3 cm) or double (5 cm) applications of shotcrete are


applied - - usually in combination with systematic bolting (i. e. support
categories 21 and 25, Tables 12 and 13) - - the function of the shotcrete is
to prevent loosening, especially in the zone between bolts. In such cases no
attempt was made to use Eq. (11) for design thicknesses. The mode of failure
of thin layers of shotcrete is one of shear, not bending or compression, as
emphasised by R a b c e w i c z (1969) and M i i l l e r (1970). In fact, the support
tables are based on a wealth of case records in these support categories, and
any attempt to incorporate theory would be superfluous, even if the relevant
theory was reliable.
234 N. Barton, R. Lien, and J. L u n d e :

Conclusions
1. The method of classifying rock masses for tunneling stability incor-
porates six parameters which can be estimated in the first instance using an
inexpensive combination of field mapping and geological engineering judge-
ment. Should bore core be available together with the results of rock mechan-
ics tests such as point load strength, natural rock stress etc., then the esti-
mate of rock mass quality Q will be that much more reliable, though not
necessarily more accurate. At a more advanced stage of a project when
exploratory adits are available, the estimates of Q can, and should be updated
further. Support requirements may be re-evaluated in the light of the in situ
conditions revealed.
2. The support recommendations contained in this paper are based for
the most part on general engineering practice for a given type of excavation.
If for some reason the quality of drilling and blasting is better or worse than
that in normal practice, then the recommended support will tend to be
over-conservative or inadequate respectively. The most reliable estimates of
rock mass quality Q and support measures will therefore be obtained from
exposures excavated by the same methods as those to be used in the final
excavations. Where this is not possible, allowance should be made, partic-
ularly with regard to the value of J,~ (joint set number) and RQD.
3. The use of past and present case records as a basis for future design
introduces the danger of perpetuating over-conservative (and occasionally
under-conservative) practice. Consequently, case records describing failure
of temporary support, or the necessity for additional support are especially
valuable for indicating what the present safety margins are.
4. A further danger of using past and present case records as a basis for
future design is that excavation techniques are changing. More and more
smooth wall blasting is used and more and more tunnels are machinebored.
The support pressure required will reduce as improved excavation techniques
result in less disturbance of the surrounding rock. These trends must be
incorporated as they occur.
5. Readers in a position to supply detailed case records, especially in
areas where the authors' data is sparse, could make a valuable contribution,
enabling the updating and improvement of the support tables.

Acknowledgements :
It is not usual to acknowledge the contribution of publications. How-
ever, the field work performed by C e c i l (1970) has proved such a valuable
source of information for developing this method of classification that his
contribution must be specially acknowledge. The review article by C o r d i n g,
H e n d r o n and D e e r e (1972) was another valuable source from the Univer-
sity of Illinois. Finally the authors would like to thank their colleagues at
the Norwegian Geotechnical Institute, in particular Arild P a l m s t r6 m, Tor
L 6 k e n , Tore V a l s t a d and Bj6rn B u e n for helpful discussisons and con-
tributions.
Engineering Classification of Rock Masses for the Design of Tunnel Support 235

References

B a r t h , S.: Felsmechanische Probleme beim Entwurf der Kaverne des Pump-


speicherwerkes Waldeck II. Bautechnik, Vol. 49, No. 3, 73--83 (1972).
B a r t o n , N.: Review of a new shear-strength criterion for rock joints. To
be published in: Engineering Geology, Amst., Vol. 7 (4), 287--332 (1973).
B a r t o n , N.: A review of the shear strength of gilled discontinuities in rock.
Norsk forening for fjellsprengningsteknikk. Fjellsprengningsteknikk, Bergmekanikk
(Conference), Oslo 1973. Tapir, Trondheim. Editor: E. B r o c h . 38 p. (1974).
B a r t o n , N., R. Lien, and J. L u n d e : Analysis of rock mass quality and sup-
port practice in tunneling, and a guide for estimating support requirements. NGI
Internal report 54206, 74 p. (1974).
B e n s o n , R. P., R. J. C o n l o n , A. H. M e r r i t t , P. J o l i - C o e u r , and D. U.
D e e r e : Rock mechanics at Churchill Falls. American Society of Civil Engineers.
Symposium on Underground Rock Chambers, Phoenix, Arizona. Proceedings,
407--486 (1971).
B i e n i a w s k i , Z.T.: Engineering classification of jointed rock masses. The
Civil Engineer in South Africa, Dec. 335--343 (1973).
B j e r r u m , L.: Discussion on: R.E. G o o d m a n and H . M . E w o l d s o n : A
design approach for rock bolt reinforcement in underground galleries. International
Symposium on Large Permanent Underground Openings. Oslo 1969. Proceedings,
p. 261 (1970).
B r o w n, G. L., E. D. M o r g a n, and J. S. D o d d : Rock stabilization at Mor-
row Point power plant, American Society of Civil Engineers. Proceedings, Vol. 97,
No. SM 1, 119--139 (1971).
C e c i l III, O. S.: Correlations of rock bolt-shotcrete support and rock quality
parameters in Scandinavian tunnels. Ph. D. Thesis. Urbana, University of Illinois.
414 p. (1970).
C o a t e s , D. F.: Classification of rocks for rock mechanics. Rock Mechanics
and Mining Sciences, Vol. 1, 421--429 (1964).
C o r d i n g , E.J., and D.U. D e e r e : Rock tunnel supports and field mea-
surements. North American Rapid Excavation and Tunneling Conference, Chicago
1972. Proceedings, Vol. 1, 567--600 (1972).
C o r d i n g , E.J., A.J. H e n d r o n , and D.U. D e e r e : Rock engineering for
underground caverns. American Society of Civil Engineers. Symposium on Under-
ground Chambers. Phoenix, Arizona, 1971, 567--600 (1972).
D e e r e , D. U.: Technical description of rock cores for engineering purposes.
Felsmechanik und Ingenieurgeologie, Vol. 1, No. 1, 16--22 (1963).
D e e r e , D . U . , R . B . Peck, J.E. M o n s e e s , H.W. P a r k e r , andB. S c h m i d t :
Design of tunnel support systems. Proceedings. 49th Annual Meeting, Highway
Research Board, January (1970).
D e n k h a u s , H.G.: Discussion on: Theme I: Engineering geological con-
siderations in the design and construction of large underground openings. Inter-
national Symposium on Large Permanent Underground Openings. Oslo 1969. Pro-
ceedings, 125--126 (1970).
M e r r it t, A. H.: Geological predictions for underground excavations. North
American Rapid Excavations and Tunneling Conference, Chicago 1972. Pro-
ceedings, Vol. 1, 115--132 (1972).
236 N. B a r t o n et al.: Engineering Classification of Rock Masses

M o n s e e s , J.: Design of support systems for tunnels in rock. Ph.D. Thesis.


University of Illinois. 1--252 (1970).
M/.i I 1e r- Salzburg, L.: A new European tunneling concept. Paper presented at
a Tunneling Conference at Lorch, West-Germany, 1970, under the title "Neuere
Auffassungen im mitteleurop~iischen Felshohlraumbau und deren Auswirkungen
auf die Praxis". Salzburg. 42 p. Osterreichische Gesellschaft fi.ir Geomechanik.
Translation 17 (1970).
P a l m s t r 6 m , A.: Private communication (1974).
R a b c e w i c z , L.v.: Stability of tunnels under rock load. Water Power, Vol. 21,
June: 225--229, July: 266--273, August: 297--302 (1969).
S e l m e r - O l s e n , R.: Experiences with using bolts and shotcrete in area with
rock bursting phenomena. International Symposium on Large Permanent Under-
ground Openings, Oslo 1969. Proceedings, 275--278 (1970).
T e r z a g h i , K.: Rock defects and loads on tunnel supports. P r o c t o r , R.V.,
and T. L. W h i t e : Rock tunneling with steel supports. Youngstown, Ohio, Com-
mercial Shearing and Stamping Co., 17--99. Harvard University. Graduate School
of Engineering. Publication, 418 - - Soil mechanics series 25 (1946).
W i e k h a m , G. E., H. R. T i e d e m a n n , and E. H. S k i n n e r : Support deter-
minations based on geological predictions. North American Rapid Excavation
and Tunneling Conference. Chicago 1972. Proceedings, Vol. 1, 43--64 (1972).

Address of the authors: Dr. Nick B a r t on, Norwegian Geotechnical Institute,


Postboks 40 Tasen, Oslo 8, Norway.